438 THE ETSIN-GOL DELTA AND THE RUINS OF KHARA-KHOTO [Chap.XIII
for eight days, and was interrupted only for the examination of the ruined dwellings that Afrâzgul succeeded in discovering at some distance eastwards. Its difficulties were considerably increased by the innate laziness of the Mongols in their unaccustomed role as labourers, and no less by the quarrels that our irascible ` Lama' Malum managed to pick with this refractory crew. In describing the résults of our labours I shall ignore the order in which they were carried out and deal with the remains examined in quasi-topographical order.'
The most striking ruins of Khara-khoto are those of its circumvallation. This forms, as the sketch-plan in Pl. 18 shows, an approximate rectangle nearly orientated. The area enclosed measures about 466 yards on the north side and 381 yards on the west. It is thus about 5o per cent. larger than that of the Lou-lan station,' but less than one-half that of the ruined town of So-yangch`êng near Ch`iao-tzû.3 The walls are built of stamped clay and reinforced by a wooden framework of which the big rafters could be traced in three rows all round the inside faces of the walls. But in most places their position is marked only by the holes which the decayed timber has left (Fig. 248). The walls are about 38 feet thick at the base, but show a considerable inward slope so that the width at the top, about 3o feet from the ground, is only 12 feet. This width is, however, greatly increased near the north-western corner (Fig. 248), where the top is crowned by the Stûpas to be presently mentioned, with a correspondingly greater thickness at the base. In places a parapet about one foot thick, with loopholes, still survives to a height of 5 or 6 feet. Ramps leading up to the top of the walls can be traced at the gates and at the north-western and south-eastern corners.
Gates, 18 feet wide, lead through the western and eastern wall faces, each protected by a rectangular outwork built as massively as the walls themselves (Fig. 25o). In addition, the walls have been pierced at two points (z, 2 in Pl. i8) by passages which are certainly of later date. The one in the south wall appeared to me as if made to facilitate access to the ground outside at some period when defence had ceased to be a serious consideration. The other in the north wall (Fig. 247) seemed needlessly large for this purpose. The fact that both are quite close by the side of a bastion clearly militates against the theory of a breach made by a besieger. On the other hand, I found no evidence to support the story about the cutting in the north wall heard by Colonel Kozlov, which
1 [It may conveniently be explained here that at the time of my visit to Khara-khoto the only account available to me of the explorations which Colonel Kozlov had carried out at the site in 1908 was that contained in the English translation of his preliminary reports to the Imperial Russian Society of Geography of his journey, 1907-9, as published in the Geographical Journal, xxxiv. pp. 384-408 (October 1909), and xxxvi. pp. 288-310 (September 1910). Nor had I access to any more detailed description of the ruins as the distinguished Russian explorer had found them or of the operations to which he devoted his two successive stays at the site, when I was preparing the present record of my own work there.
It is only at the time when this chapter is passing into print that I received from Colonel Kozlov himself a copy of his book dealing with ` Mongolia, Amdo and the dead city of Khara-khoto ' (in Russian ; Petrograd, 1923). My regrettable ignorance of Russian does not allow me definitely to ascertain what additional information, if any, of archaeological import Chapter xxv dealing with the site (pp. 546-68, with 23 figs. in the text) furnishes. It contains, however, very welcome reproductions of photographs taken of the
sepulchral monument (marked by me K.K. it ; see below) from which Colonel Kozlov secured his great trouvaille of Buddhist texts and paintings.
These photographs had already before been published by M. Serge d'Oldenburg in the introductory chapter of his monograph on ` Materials for Buddhist Iconography from Khara-khoto ' (in Russian ; St. Petersburg, 2924). This valuable publication from the pen of a very competent expert deals with selected specimens of the many fine paintings discovered by Colonel Kozlov. It too, I regret, remained unknown to me until I received it in May 1925 as well as the first-named book as kind gifts from the authors. Extracts from the iconographic information furnished in M. d'Oldenburg's monograph appear to constitute the bulk of the remaining chapters, XXVI—VIII, devoted to Kharakhoto in Colonel Kozlov's volume.
Notices on Khara-khoto, obviously of a preliminary character, by MM. A. Ivanov, d'Oldenburg, V. L. Kotwitz, are quoted in a foot-note of M. d'Oldenburg's monograph, p. I, from the Proceedings of the Imperial Russian Geographical Society (vol. xlv, 1909, pp. 463-77).]
2 See Serindia, i. p. 387. 3 See ibid., iii. p. 1102.